During adolescence, individuals often receive their first exposure to alcohol, and a significant proportion do so during episodes of high intake or bingeing. Such experience can be antecedent to problem drinking and is associated with impairments in decision making. Recently, it has been demonstrated that adolescent alcohol use is sufficient to produce long-term perturbation of risk-based decision making in rodents. Adolescence is a critical period of maturation where brain development may be disrupted by alcohol use. Specifically, the mesolimbic dopamine system has been shown to be enduringly altered by chronic alcohol exposure during adolescence. Phasic increases in dopamine are evoked by rewarding outcomes and associated cues. These signals have been shown to scale with the magnitude and probability of reward which, together with reward costs, make up the fundamental components of optimal decision making. Importantly, all of these attributes of the decision making apparatus are thought to be exploited by abused substances. Indeed, phasic dopamine signaling to risky, but not safe, options is increased in animals with a history of adolescent alcohol exposure. These findings suggest that changes in phasic dopamine release, as a consequence of alcohol exposure, could underlie biases in choice behavior and promote risk preference. Therefore, it is hypothesized that adolescent alcohol exposure influences risk preference through modulation of dopamine systems and that such modulation perturbs one or more of three fundamental elements of decision making; cost encoding, risk assessment, and/or the encoding of reward outcomes during learning. The current proposal will test these hypotheses with three specific aims. Aim 1 will test the hypothesis that risk preference evolves from cost insensitivity where anticipated reward value is not discounted based upon the increasing cost associated with its procurement. Aim 2 will test the hypothesis that risk preference results from diminished risk assessment and the proposition that uncertainty may paradoxically enhance value (i.e. a gambling buzz). Aim 3 will test the hypothesis that risk preference is a consequence of the aberrant encoding of reward outcomes during reinforcement learning.